Constructive Empiricism and Interpretation
For thirty-some years now, the philosophy Bas van Fraassen has been defending a position in the philosophy of science known as constructive empiricism. It is a nuanced view, but it is characterized by a central theme: making empiricism plausible. Empiricism for van Fraassen is a “stance”—an attitude with no content itself (empirical or metaphysical), but which determines how one responds to particular bits of content. Contrasting with empiricism is the stance van Fraassen calls materialism. Van Fraassen characterizes the two stances as follows (The Empirical Stance, 62-63):
Characteristic of materialism is a certain deference to the content of science. This deference takes two forms: the belief that the scientific description of the world is true, in its entirety or near enough, at least a strong inclination toward completeness claims for the content of certain sciences. …
Empiricism may also be approached through reflection on its positive attitude toward science. But this admiring attitude is not directed so much to the content of the sciences as to their forms and practices of inquiry. Science is a paradigm of rational inquiry. To take it as such is precisely to take up one of the most central attitudes in the empiricist stance. But one may take it so while showing little deference to the content of any science per se. …
For the materialist, science is what teaches us what to believe. For the empiricist, science is more nearly what teaches us how to give up our beliefs. All our factual beliefs are to be given over as hostages to fortune, to the fortunes of future empirical evidence, and given up when they fail, without succumbing to despair, cynicism, or debilitating relativism.
Importantly, van Fraassen does not think that one stance is more rational than the other (though he thinks many actual materialists go astray and endorse what he considers “inflationary metaphysics”—e.g. modal realism). The constructive empiricist believes that successful scientific theories are empirically adequate (which means that they save all of the observable phenomena), but not that they are strictly true, which would require that they be true about the unobservable phenomena as well. For the constructive empiricist, all that is required for scientific inquiry is that, when one accepts a scientific theory, the only belief to which one is committed is that the theory is empirically adequate. Van Fraassen adds that one is also committed to further testing of the theory, and to giving it up if it should prove not to be empirically adequate. Importantly, however, van Fraassen does not believe that it is irrational to believe that the best scientific theories are true; he simply believes it is not rationally compelled. Neither the empiricist nor the materialist can accuse the other of irrationality.
It is important to recognize that constructive empiricism does not involve any ontological claims about the unobservable. What is observable, for the constructive empiricist, is what is observable-for-humans, as our best current theories of perception delimit what we can observe. But it would be quite strange to believe that nothing that exists escapes detection by our perceptual systems. The constructive empiricist only abjures accepting any claims about what it is that so escapes our perceptual faculties.
How might one recognize an empiricist or a materialist? The materialist is more likely, for one thing, to think that ethical and aesthetic judgments amount to nothing more than opinion, for there is no hope of deriving e.g. an ethical philosophy from the content of scientific theories. The empiricist is more likely to think that ethics, interpretation, aesthetics, etc., are matters of serious rational inquiry. Note that in the quote above, van Fraassen describes the empiricist as believing that science is a paradigm of rational inquiry, but certainly not the paradigm.
The dichotomy between these two stances is not exhaustive. Also possible are stances that do not think science is a paradigm of rational inquiry at all (though this is much more difficult to maintain sensibly these days), and, more pertinently to my case, stances in between materialism and empiricism. I am sympathetic to aspects of both views. The empiricist restriction of belief to beliefs about observable phenomena is not something I find particularly plausible. Nevertheless, I am sympathetic to van Fraassen’s modal nominalism, and I certainly side with the empiricist in thinking that rational inquiry extends well beyond the rather rigid confines of scientific inquiry. Take what David Foster Wallace once said in an interview:
In dark times, the definition of good art would seem to be art that locates and applies CPR to those elements of what’s human and magical that still live and glow despite the times’ darkness. Really good fiction could have as dark a worldview as it wished, but it’d find a way both to depict this world and to illuminate the possibilities for being alive and human in it.
I think what David Foster Wallace is doing is setting out a task for a type of inquiry carried out by artists. This inquiry requires several steps: the artist must first locate what is “human and magical” in the world, and then she must apply CPR to these elements. The CPR metaphor is apt: the artist not only illuminates “the possibilities for being alive and human” at a particular time, the artist brings these possibilities back to life. The artist does this through some marriage of form and content. One need not think that there are determinate answers to the question “what are the possibilities for being alive and human in today’s world” in the way there is a determinate answer to the question, “what is the charge of an electron?” in order to think that this is a domain of inquiry that requires exacting rigor and the ability to give reasons that have more than a mere subjective validity. Does art discover these possibilities or create them? The question is really immaterial. Good art shows, on a rigorous and I would contend rational basis, what possibilities there are. (I should add that I do not intend to imply that this is the only task of art. But it is I think indubitably a task.)
But my topic here is neither science nor the creation of art, for I am neither a scientist nor an artist. What I engage in with this blog is a mixture of philosophy and interpretation, and in this case, philosophy of interpretation. However one feels about constructive empiricism as a philosophy of science, I think it is particularly useful in thinking about interpretation. I will try to elucidate this thought at an abstract level, providing examples where I can. Before I begin, let me be clear that I don’t know if I agree with the view I will sketch out. This is primarily an attempt to sketch out a position, with the hopes that being forced to formulate it explicitly will help me to evaluate it going forward. So my acceptance of what I say below is provisional. With that, onward.
Van Fraassen’s claim that acceptance of a scientific theory does not require acceptance of its claims about unobservable entities is notorious, but an analog of this claim in the case of interpretation seems to me like it at least ought to be uncontentious. Here goes: an interpretation of a work of art must be consistent with everything contained within the work (i.e. it must save the “observable phenomena” regarding the content of the work), but its claims about what is not part of the content of the text need not be literally true (of the work).
What does this mean? In a film, certain events are shown on the screen, but only rarely is a film so purely self-contained that everything that matters occurs onscreen. There are time gaps between scenes. There are times when characters wander in and out of frame in the middle of a scene. Interpretation of the film is probably nigh on impossible if the interpreter cannot give some account of what is going on in instances like these. What is observable in a film, however, is nothing more or less than what is seen and heard while the film is playing. Any interpretation must be consistent with what is seen or heard.
At this point there is a disanalogy between the constructive empiricist approach to interpretation and the constructive empiricist approach to science. Regarding, say, physics, the constructive empiricist is perfectly happy to admit that there might be electrons—she simply does not feel compelled to believe there are. But in the case of a film, the literally true answer to what happens off-screen is: nothing. It is perfectly appropriate to think of the film as creating a world that encompasses more than just what is seen and heard in watching the film, but it is a mistake to reify this world. What exists is nothing more or less than what is contained in the work of art. It is a strange ontology that limits what exists in our “physical” world to what we can perceive, but in a work of art this ontological restriction is exactly right. What exists is what is shown (film, painting), heard (film, music), or described (literature).
Nevertheless, in interpretation, it may be absolutely crucial to make claims about what happens that are not specified in the work itself. In order to “save the phenomena” of a work, it may be necessary to have an account that goes beyond these phenomena. And this will involve a sort of talking as if those events are every bit as real as what is contained within the work. There is a substantial gap in time between where Infinite Jest ends and where it begins, and any interpretation of the work must say something about what occurs to bridge this gap, but there can be no literally true answer. One cannot save the phenomena of Infinite Jest while remaining silent about this gap.
There is another important analogy between scientific and interpretive constructive empiricism at this point. In constructive empiricism about science, to save the phenomena is to be consistent with them in a logical sense. Any consistent theory will suffice as empirically adequate. Nevertheless, we want scientific theories to do more than be empirically adequate; in particular, we want them to explain. Van Fraassen devotes a chapter of The Scientific Image (his first book-length defense of constructive empiricism) to arguing that explanation is a pragmatic virtue of theories, and does not provide any indication to think that they are true. It is a valid reason for acceptance in the constructive empiricist sense, but explanatory power does not compel a realist attitude. This demand for explanatory power is at least one reason why science cannot consist of just an exhaustive list of patterns among the phenomena. Likewise, an interpretation of a work of art cannot just be a description of what occurs in the work: it must make these occurrences sensible. And just as this may require the positing of unobservable entities like electrons in the case of physical theory, so may interpretation of a work of art require positing “unobservable” events.
One final point. Nelson Goodman in Ways of Worldmaking suggests that ‘true’ and ‘false’ (conceived, I think, along the lines of van Fraassen’s “literally true” and “literally false”) are not exhaustive of the ways we may be right or wrong. To these limited notions Goodman would add notions of metaphorical truth and falsity, as well as a general purpose sense of rightness and wrongness that can be expanded to include anything appropriate—no doubt Goodman himself favors inclusiveness in fleshing out just how things may be right or wrong. A description of what happens in a text may be literally true or false, but interpretation goes well beyond description, and so is better evaluated as right or wrong. And so, even though claims about what happens off-screen or outside the text or beyond the picture frame may be literally false (because what is beyond the work of art literally does not exist), they are nonetheless subject to standards of right and wrong. If I posit that, during the time between the events that end Infinite Jest and those that begin it, Descartes’ evil demon freezes time, moves everybody five inches to their right, and then unfreezes time, I am wrong, despite my claim being no less literally false than any other that describes that time. And while interpretation may not generally be as clear-cut as that (indeed, no interesting interpretive claim will be), it is still subject to the need for rigorous reasoning.